How thick does closed-cell foam have to be before it is the vapour barrier?

Two supporting articles complete the chain. Article 9.25.4.1.(1) establishes that thermally insulated wall, ceiling and floor assemblies need a vapour barrier at all, to stop diffusion from the interior into wall spaces, floor spaces and attic or roof spaces. And Sentence 9.25.4.2.(2) carves out thermally insulated foundation wall assemblies, which may use a variable-permeance “smart” barrier: ≤ 60 ng/(Pa·s·m²) dry cup and > 300 ng/(Pa·s·m²) wet cup. NBC(AE) · 9.25.4.1.(1), 9.25.4.2.(2)

These are published values from CCMC evaluations and manufacturers’ Canadian technical data. They are not offered as a product recommendation — they are here so that a reader can see how much headroom a conforming product has against the 60 ng limit.

1. Many products also pass at 25 mm — but that is a product result, not a code entitlement. Insulthane Extreme measures 47 ng at 25 mm; a legacy WALLTITE formulation measured 51. Both are under 60. The material standard only guarantees the threshold at 50 mm. The correct sentence is: “50 mm of CAN/ULC-S705.1 closed-cell foam meets the code’s vapour-barrier permeance limit; some products also meet it at 25 mm — check the specific CCMC report.”
2. Do not lead with 38 mm. The Spray Foam Coalition’s Canadian installation guide says most closed-cell products “typically qualify as a code-compliant vapour barrier at 38 mm thickness or greater.” That is an industry generalisation from a trade association, not a code number and not a standard’s test thickness. It may well be true of a given product; it is not a specification. ACC / SFC Canadian guidesec.
3. CCMC permeance is measured with the top skin removed — so field foam is typically tighter than the certificate. The Soprema evaluation states it: “The WVP is determined from a core sample with the top skin removed. Due to the effect of the skin, the WVP at this thickness may be lower in the site-installed product.” The certificate is the conservative number.
4. Watch the unit conversion. 1 US perm = 57.4 ng/(Pa·s·m²). The code number is 60 ng, not 57.4. They are close but not identical — RDH’s Alberta report, for instance, works to “less than 57 ng/Pa·s·m² or approximately 1 US Perm.” When you are quoting the code, quote 60 ng.

Design conditions in Red Deer means a January 2.5% temperature of −32 °C and a 1% temperature of −35 °C. That is the outdoor number the condensation check is run against — one more reason the Zone 7A design data is worth having in front of you before specifying a thickness.

A related objection this resolves: “what about the rafters?” Wood framing already meets the vapour barrier requirement on its own. RDH cites Exova test data putting 2×4 framing at 36 ng/(Pa·s·m²) (0.63 US perm) and 2×6 framing at 30 ng (0.52 US perm), dry cup — both comfortably inside the 60 ng limit. A full-depth foam roof does not have a vapour hole where each rafter is. RDH, citing Exova 09-06-M0097

Subsection 9.25.5 governs low-permeance materials placed outboard of the insulation, and in Zone 7A it has teeth. Article 9.25.5.1.(1) catches sheet and panel-type materials with an air leakage characteristic under 0.1 L/(s·m²) at 75 Pa and a water vapour permeance under 60 ng/(Pa·s·m²) dry cup. Article 9.25.5.2.(1) then requires such materials to be installed on the warm face, or outboard of an air space vented to the outdoors, or with a minimum ratio of outboard to inboard thermal resistance taken from Table 9.25.5.2. NBC(AE) · 9.25.5.1, 9.25.5.2

Translated for Red Deer: if closed-cell foam or foam board goes on the outside of a wall, at least 30 per cent of the assembly’s total thermal resistance has to sit outboard of the foam’s inner face — otherwise that face runs cold enough to condense. On a 2×6 batt wall with roughly RSI 3.0 inboard, that is about RSI 0.9 (R-5) of continuous exterior foam as a floor, not a token 12 mm. Getting this wrong builds a condensing plane inside a wall.

Every item here is publicly checkable, which is the point. A proposal that claims foam is doing vapour-barrier duty should survive all six.

1. The product’s CCMC number, and that its status reads Active on the NRC site. The look-up pattern is public: nrc.canada.ca/…/document.html?id=NUMBER&type=cert.
2. That the report lists compliance with NBC 2020, not only NBC 2015.
3. The 50 mm water vapour permeance on the report — under 60 ng/(Pa·s·m²) if the foam is the vapour barrier.
4. The 50 mm design LTTR, and that it matches the R-value quoted on the proposal. In Canada the aged LTTR is the design value by rule of the standard, so a proposal quoting an initial ASTM C518 figure is quoting the wrong number.
5. The site quality assurance provider named in the report, and that the crew’s cards match it. There are four CCMC-recognised programmes — CUFCA, Caliber, BPC and UFC — and each manufacturer designates one. Alberta’s code names none of them; it names CAN/ULC-S705.2, through Article 9.25.2.5.
6. The Daily Work Record for the job, and the thermal barrier specified for every surface where foam will be visible.